I'm trying to design a simple device that has the same effect as using helium to distort someone's voice.
My idea is simple. Simply shift the spectrum of the input sound with a constant frequency.
voice in -> amplify -> mixer -> amplify -> output
^
|
fosc
I haven't decided which frequency fosc should have but I assume 2k-4kHz would do it.
So I started out with the mixer as I assumed that this would be the most difficult. I read that some diodes approximate V=aI^2 and figured that I could use the diode as mixer device. I put my signal at 50mVpp 10kHz on top of a 0.65mV offset and put that voltage over a normal diode. The negative side of the diode is connected to a transimpediance amplifier built with a simple opamp. The problem is that I don't see any kind of sign of the higher frequencies at the output signal. I do see a slightly distorted sine wave at the input frequency. I have tried filtering out the second harmonic with a rather steep filter but I still don't see what I expect on the output. I also tried to simulate the circuit in spice just to verify that the theory works. In spice I saw a second harmonic of about 40% amlitude compared to the input signal.
Does anyone know where I go wrong? Can the mixer not be implemented this way using a normal diode? Maybe the diode doesn't approximate V=aI^2? Maybe the amplitude of the second harmonic will be too small to filter out?
If anyone has any other ideas of how I could build a 'helim emulator'
I would be glad to hear them.
My idea is simple. Simply shift the spectrum of the input sound with a constant frequency.
voice in -> amplify -> mixer -> amplify -> output
^
|
fosc
I haven't decided which frequency fosc should have but I assume 2k-4kHz would do it.
So I started out with the mixer as I assumed that this would be the most difficult. I read that some diodes approximate V=aI^2 and figured that I could use the diode as mixer device. I put my signal at 50mVpp 10kHz on top of a 0.65mV offset and put that voltage over a normal diode. The negative side of the diode is connected to a transimpediance amplifier built with a simple opamp. The problem is that I don't see any kind of sign of the higher frequencies at the output signal. I do see a slightly distorted sine wave at the input frequency. I have tried filtering out the second harmonic with a rather steep filter but I still don't see what I expect on the output. I also tried to simulate the circuit in spice just to verify that the theory works. In spice I saw a second harmonic of about 40% amlitude compared to the input signal.
Does anyone know where I go wrong? Can the mixer not be implemented this way using a normal diode? Maybe the diode doesn't approximate V=aI^2? Maybe the amplitude of the second harmonic will be too small to filter out?
If anyone has any other ideas of how I could build a 'helim emulator'
I would be glad to hear them.
you have a computer. there are free programs which can be a VST host and free vst plugins that can pitch shift. (if you want real time)
is there a reason it has to be a electronic device?
i think there are toy "voice disguisers" that can also do that.
is there a reason it has to be a electronic device?
i think there are toy "voice disguisers" that can also do that.
Hi,
If you don’t need real time http://www.goldwave.com/ is a nice program that has pitch shift and mechanise effects. It is full functional for free.
😉
If you don’t need real time http://www.goldwave.com/ is a nice program that has pitch shift and mechanise effects. It is full functional for free.
😉
Great, I found several schematics these.Miniwatt said:Use a pitch-shifter
No, I'm just doing it for learning electronis.neutron7 said:
Thanks for all your replies.
A ringmodulator transforms two input frequencies to their sum and their difference frequncies. You could use your voice input and a fixed frequency to shift the voice (much lower than your frequencies). For a single output you have to use complex signals (hilbert transformation, two allpass filters to get constant 90deg phase difference) for ringmodulation. Look for a possible schematic of the famous BODE pitch or frequency shifter, but this is no trivial stuff.
A simple ringmodulator will probably do if one of the output signals is far off.
Just by googling you will find lots of schematics.
The simplest passive ringmodulator looks like that:
http://www.synthdiy.com/files/1/200248/ringmod.jpg
This ringmodulator needs no external power.
Two (2) Stancor A-4713 driver transformer or equivalent
Four (4) diodes GE45A or equivalent
Three (3) suitable jacks
One (1) mounting box
Many small transformers will work to some degree.
The diodes must be germanium and not silicon types.
A little more sophisticated:
http://www.uni-bonn.de/~uzs159/ringmod.png
http://www.analog-synth.de/synths/ringmod/digital_ringmodulator.htm
A simple ringmodulator will probably do if one of the output signals is far off.
Just by googling you will find lots of schematics.
The simplest passive ringmodulator looks like that:
http://www.synthdiy.com/files/1/200248/ringmod.jpg
This ringmodulator needs no external power.
Two (2) Stancor A-4713 driver transformer or equivalent
Four (4) diodes GE45A or equivalent
Three (3) suitable jacks
One (1) mounting box
Many small transformers will work to some degree.
The diodes must be germanium and not silicon types.
A little more sophisticated:
http://www.uni-bonn.de/~uzs159/ringmod.png
http://www.analog-synth.de/synths/ringmod/digital_ringmodulator.htm
- Status
- Not open for further replies.